The data presented in this study were collected as part of a prospective, multicentre, randomized clinical trial (The Cardiovascular Health/Outcomes: Improvements Created by Exercise and education in SCI; CHOICES). This clinical trial was pre-registered (NCT01718977) and a detailed protocol manuscript has previously been published.15 Continuously measured cardiovascular control data were only collected at one study site (International Collaboration of Repair Discoveries in the Blusson Spinal Cord Centre, Vancouver, BC, Canada) and are reported in this manuscript. This protocol was approved by the University of British Columbia Clinical Research Ethics Board (H12-02945) and all participants provided written informed consent prior to enrolling in the trial. Participants were randomly allocated (1:1) to either 6 months of ACET or BWSTT using a central, web-based computer randomization service (Empower health research inc.).
Participants
Individuals were eligible for inclusion in this clinical trial if they were aged between 18 – 60 years, with a traumatic, chronic (≥1-year post-injury), motor-complete SCI (American Spinal Injury Association Impairment Scale [AIS] A or B) between the fourth cervical and sixth thoracic spinal cord segments (C4–T6). SCI severity and lesion level were determined by the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI; 2011 version)16 exam and performed by a trained physician (A.V.K). Furthermore, eligible participants had a carotid-to-femoral pulse wave velocity (CHOICES primary outcome measure) ≥ the norm median value for age-matched uninjured individuals,17 and were thus considered to have an established CVD risk profile. Exclusion criteria were as follows: 1) any active or historical medical issues (e.g. cardiovascular or cardiopulmonary problems/disease) that preclude the enrolment in an exercise intervention, 2) a body mass in excess of 135 kg (absolute capacity of the body-weight supported treadmill).
Exercise interventions
Participants performed three sessions per week, ultimately completing continuous exercise (30 minutes per session for ACET or 60 minutes per session for BWSTT) over a period of at least 6 months. Importantly, ACET was prescribed according to the updated SCI-specific exercise guidelines to improve cardiometabolic health, (3 x 30 mins per week)6 with participants performing moderate-to-vigorous intensity aerobic exercise with a rating of perceived exertion (RPE) between 11 and 16. At present, there are no accepted guidelines for the optimal BWSTT stimulus, therefore, 60 minutes was chosen based on prior BWSTT interventions11,18 and to approximate training loads between the two groups due to the nature of passive lower extremity exercise. ACET was performed with a cadence set at >50 revolutions per minute on a wall mounted arm-crank ergometer (Lode BV, Groningen, The Netherlands) at an individually determined level of resistance, which increased over the trial duration to ensure an element of progression. BWSTT was performed on a treadmill (Woodway, Weil am Rhein, Germany) using a body weight-support system (Andago, Hocoma AG, Volketswil, Switzerland) with the assistance of at least two trainers. Considering the associated medical and non-medical complexities of working with this population, a pragmatic approach was adopted whereby participants were given the opportunity to compensate for missed exercise sessions, ensuring they completed 72 exercise sessions in total. The adherence criteria and detailed assessment method is described in the protocol manuscript.15
Outcome measures
This trial addresses changes in cardiovascular control using laboratory assessed hemodynamics responses to orthostatic stress, spontaneous cvBRS, and HR/blood pressure (BP) variability, which were secondary outcomes of the broader CHOICES clinical trial.15 Data regarding changes in the primary outcome measure (arterial stiffness) and other secondary outcomes (cardiorespiratory fitness and cardiometabolic health) are published elsewhere.19 Cardiovascular control outcomes were assessed at baseline and after cessation of the exercise interventions (~6 months), using the methods outlined below.
Orthostatic Intolerance: Laboratory Sit Up Test and Hemodynamics
Participant hemodynamic responses and susceptibility to orthostatic hypotension (OH) was assessed at baseline and 6 months using a standardised and reliable sit-up test (SUT).20,21 Prior to testing, participants were asked to: 1) avoid performing strenuous physical activity (>24 hrs), 2) abstain from ingesting caffeine, nicotine or alcohol (>12 hours), 3) fast (>4 hrs) and, 4) empty their bladder prior to testing to minimise the risk of reflex sympathetic activation. Participants were passively moved to an upright seated position for 15-min, following 10-min of supine rest. HR and beat-by-beat BP were measured continuously via single-lead ECG (Model ML132; ADInstruments, Colorado Springs, Colorado, USA) and by a non-invasive BP device that places a cuff on the right middle finger (Finometer PRO; Finapres Medical Systems, Amsterdam, The Netherlands), respectively. BP was also measured at 1-min intervals using an automated device (Dinamap Carescape V100; GE Healthcare, Buckinghamshire, UK) with a cuff placed on the left upper arm and used to correct for minor deviations in height between the heart and cuff measurements from the finger. Cautious monitoring of the participant ensured that any severe symptoms of presyncope or autonomic dysreflexia resulted in the termination of the test. In addition, stroke volume (SV), cardiac output (Q), and total peripheral resistance (TPR) were derived from the raw finger BP waveforms using the previously used and validated Modelflow® method.22,23
Heart Rate and Systolic Blood Pressure Variability
HR and BP variability were collected at baseline as well as during SUT and used as surrogate markers of autonomic cardiovascular control.24,25 ECG filtering and R-wave detection was performed offline using LabChart (V8, ADInstruments, Colorado Springs, Colorado, USA). Time and frequency domain analysis of the RR-interval time series was performed in accordance with the European task force HRV guidelines24 and described elsewhere.26,27 Time domain parameters included the standard deviation of all normal RR-intervals (SDNN), the root mean squared of successive differences in RR-interval (RMSSD), and percentage of RR-intervals longer than 50ms of the previous RR-interval (pNN50). Power spectral density of the RR-interval spectrum was calculated using Welsh’s method with 50% windowing. Low frequency (LF; 0.4-1.5 Hz) and high frequency (HF; 1.5-5 Hz) powers were then extracted from the total power spectral density curve and normalized as a percentage of total power. LF and HF systolic BP variability (BPV), measures of sympathetic vasomotor tone,28 were assessed using the same methods outlined for HRV.
Spontaneous Cardiovagal Baroreflex Sensitivity
Spontaneous cvBRS was assessed using the sequence method, a valid and reliable measure of cvBRS compared to pharmacological assessments.29,30 A custom Excel spreadsheet (V16.48, Microsoft, Redmond, Washington, USA) was used to detect sequences of three or more consecutive beats in which both SBP and RR-interval of the same beat progressively increase (up-sequence) or decrease (down-sequence). The minimum beat-by-beat changes required were at least 1 mmHg for SBP and at least 1.0 ms for RR interval.30,31 Simple linear regression was used to relate SBP to RR-interval for each sequence and only sequences with an r≥0.85 were included in analysis.30,31 For each participant, the average slope for pooled sequences was taken as the individual’s cvBRS index.
Statistical Analyses
Serial measurements of episodic BP during the SUT at baseline and follow-up were converted into simple summary statistics [i.e., within-participant supine BP averaged over ten measurements, largest change (D) between average supine BP and SUT]. Responses within and between trials were analyzed by a three-way (time x position [supine vs SUT] x group) repeated measures analysis of variance (ANOVA) for raw values and a two-way (time x group) repeated measures ANOVA for D values. ANOVA’s were performed irrespective of any minor deviations from a normal distribution but with Greenhouse–Geisser corrections applied to intraindividual contrasts, where ɛ < 0.75, and the Huynh–Feldt corrections applied for less severe asphericity. Where significant interaction effects were observed, Holm-Sidak Posthoc testing was applied to determine significant differences within/between groups. Correlation analysis between cvBRS and hemodynamic/HRV parameters were assessed via Pearson correlation coefficients. All statistical analyses were performed using Statistical Package for the Social Sciences (SPSS) (version 24; IBM, Armonk, USA), with statistical significance accepted at a priori of α<0.05. Data was plotted using GraphPad Prism (version 9). Results are presented as means with upper and lower quartiles (i.e., 25-75%) or means ± standard deviations (SD) where indicated.